Summit Ventures - Light Logo
Login
Get In Touch
Anduril

Anduril
Anthropic

Anthropic
Anysphere

Anysphere
Apptronik

Apptronik
Cerebras

Cerebras
Cognition Ai

Cognition Labs
Cohere

Cohere
crusoe-squareLogo-1754598302808

Crusoe
Databricks

Databricks
Discord

Discord
Groq

Groq
kalshi

Kalshi
Kraken

Kraken
Lambda

Lambda
Lightmatter

Lightmatter
Neuralink

Neuralink
Open AI

Open AI
Perplexity

Perplexity
Polymarket

Polymarket
New_Replit_Logo

Replit
Sandbox AQ

Sandbox AQ
Saronic

Saronic
Scale Ai

Scale AI
Shield Ai

Shield AI
Space X

Space X
Stripe

Stripe

Unconventional Inc.
xAI

xAI
Home / Company / Lightmatter

Lightmatter

Last Updated November 21, 2025

Lightmatter is a photonic computing company developing optical interconnects and accelerators for large-scale AI data centers. Its technology uses silicon photonics to move data between chips with much higher bandwidth and lower energy than traditional electronic interconnects, targeting the core bottlenecks in scaling frontier AI systems. The company’s Passage platform provides 3D-stacked photonic interposers and co-packaged optics for high-density chip-to-chip connectivity, while its Envise photonic processor performs matrix operations using light. With a $4.4 billion valuation and roughly $850 million raised as of late 2024, Lightmatter is positioning itself as a foundational supplier for next-generation AI infrastructure.
Company Overview: Lightmatter
This section summarizes publicly reported factors that may be relevant to investors evaluating Lightmatter’s role in AI infrastructure. It is descriptive only and is not a recommendation to buy or sell any security. Lightmatter operates at a critical layer of the AI stack: the hardware and interconnect infrastructure required to scale large model training and inference. As model sizes and cluster scales grow, performance is increasingly constrained by bandwidth, latency, and power consumption rather than raw compute alone. By using photonics for both data movement and select compute operations, Lightmatter aims to deliver higher bandwidth and better energy efficiency than conventional electronic approaches in AI data centers. The company has raised substantial capital from institutional investors to industrialize this technology, including a $400 million Series D at a $4.4 billion valuation and total funding of about $850 million to date. These resources support manufacturing partnerships, packaging collaborations, and the ramp of products such as Passage M1000 and Passage L200. For investors focused on AI infrastructure, Lightmatter represents an example of a photonics-based approach being developed in parallel with traditional semiconductor roadmaps. At the same time, Lightmatter is a deep-tech hardware business with long development cycles, integration dependencies, and significant capital requirements. Adoption depends on hyperscalers and large data center operators incorporating photonic interconnects and accelerators into their roadmaps and workflows. Any assessment of the company must weigh these execution and adoption factors alongside the scale of the underlying AI infrastructure demand.
Investment Highlights

Valuation & Funding

  • $400M Series D closed in October 2024, led by T. Rowe Price, valuing Lightmatter at approximately $4.4B post-money.
  • Total capital raised to date is reported at roughly $850M across multiple funding rounds.
  • Series D followed a prior round less than a year earlier, reflecting increased investor focus on photonic AI infrastructure.

Commercial & Technical Traction

  • Reported 2023 revenue of roughly $50M indicates early commercial deployments beyond pure R&D, with an emphasis on AI data center use cases.
  • Partnerships with manufacturing and packaging providers such as GlobalFoundries and ASE support production readiness for Passage-based products.
  • Public demonstrations and Nature-linked research indicate photonic chips achieving precision on certain AI workloads comparable to conventional digital systems.

AI Infrastructure Focus

  • Products target bandwidth, latency, and energy constraints in large AI clusters, rather than end-user applications.
  • Passage M1000 is positioned as a 3D photonic interposer enabling 114 Tbps total optical bandwidth for next-generation AI accelerators and switches.
  • Photonic compute and interconnect offerings are aligned with growing interest in alternative architectures as traditional transistor scaling slows.

Institutional Backing

  • Backed by institutional and strategic investors including T. Rowe Price, Fidelity, and GV (Google Ventures), reflecting interest from both financial and technology ecosystems.
  • Recent funding and leadership additions (including a CFO with prior large-cap semiconductor experience) have been framed as preparation for potential future public-market options.
Product & Technology Leadership

Photonic Interconnect: Passage Platform

  • Passage™ M1000: 3D photonic “superchip” interposer announced in March 2025, providing a reported 114 Tbps total optical bandwidth to connect large arrays of accelerators and switches in a single package.
  • Passage™ L200: 3D co-packaged optics (CPO) platform designed to integrate directly with next-generation XPU and switch silicon, with 32–64+ Tbps bandwidth options to relieve interconnect bottlenecks.
  • Passage Engine: Wafer-scale photonic interconnect architecture enabling high-density chip-to-chip communication with improved bandwidth-per-watt compared to conventional electrical links.

Photonic Compute: Envise™ Accelerator

  • Envise™: General-purpose photonic AI accelerator combining electronic control and memory with photonic cores for matrix operations, targeting higher compute density and energy efficiency.
  • Photonics is used to perform linear algebra operations central to deep learning, while electronics manage data flow, memory, and system integration.

Silicon Photonics & Packaging

  • Utilizes silicon photonics processes to integrate optical components with conventional CMOS electronics.
  • Manufacturing collaborations with GlobalFoundries and advanced packaging partners (including ASE and Amkor, per public releases) to scale 3D photonic packaging and multi-die assemblies.

Software & Integration

  • Products are designed to integrate into existing AI data center workflows, with photonic components positioned as interposers, chiplets, or accelerator boards alongside established GPU/accelerator ecosystems.
  • Research and tooling efforts focus on allowing AI workloads to leverage photonic interconnect and compute without requiring complete application rewrites.
 Market Position & Strategic Advantage

Role in the AI Stack

  • Operates at the hardware and interconnect layer of AI infrastructure, focusing on how accelerators communicate and how data moves within large clusters.
  • Targets hyperscalers, cloud providers, and advanced data center operators building large-scale AI training and inference systems.

Competitive Landscape

  • Competes with both traditional semiconductor vendors adding optical features and other photonics startups pursuing similar interconnect and compute goals.
  • Positions photonics as a complementary and, in some cases, alternative path to scaling beyond the limits of purely electronic interconnects and processors.

Differentiation

  • First-mover advantage in 3D-stacked photonic interposers and co-packaged optics designed specifically for AI clusters.
  • Combination of photonic compute (Envise) and interconnect (Passage) provides a full-stack approach rather than a single-component solution.
  • Backed by institutional investors and manufacturing partners that can support scaling into large deployments if demand materializes.
Financial Opportunity

Exposure to AI Infrastructure Spending

  • Lightmatter’s business is tied to capital expenditure on AI data centers—specifically interconnect, packaging, and accelerator infrastructure.
  • If photonic interconnects and accelerators gain adoption, revenue would likely be driven by large hardware contracts and multi-year deployment programs with cloud and enterprise customers.

Business Model Characteristics

  • Hardware-centric model with potential recurring elements via follow-on deployments, maintenance, and ecosystem support.
  • High upfront R&D and capital needs, with the possibility of meaningful operating leverage if manufacturing scales successfully.
  • 2023 revenue of roughly $50M (public estimate) suggests early but non-trivial commercial traction; detailed forward revenue guidance has not been publicly disclosed.
Company Snapshot

Founded: 2017

Headquarters: Mountain View, CA, USA

Total Capital Raised: ~$850M

Latest Round: $400M Series D led by T. Rowe Price (Oct 2024)

Latest Reported Valuation: ~$4.4B post-money (Oct 2024)

2023 Revenue: ~\$50M (public third-party estimate)

Employees: ~200+ (2024, per press reports)

Primary Sector: AI infrastructure & photonic computing

Flagship Products: Passage™ M1000, Passage™ L200, Envise™ photonic accelerator

How Summit Ventures Works

Summit Ventures offers accredited investors exclusive access to shares in the secondary market, providing:

  • Network Access: Exposure to opportunities sourced through Summit’s
    relationships.
  • Market-Based Pricing: Valuations informed by current private-market activity.
  • Simplified Process: Streamlined subscription and administrative support.
  • Portfolio Exposure: Participation in select private technology companies.
Risk Disclaimer

Investment in private companies involves substantial risk and is suitable only for sophisticated investors who can bear the loss of their entire investment. Past performance is not indicative of future results.

Other Companies of Interest
View More

About Lightmatter

Lightmatter is a photonic computing company focused on removing the bandwidth, energy, and scaling bottlenecks in next-generation AI infrastructure. Founded in 2017 by MIT-affiliated researchers, the company builds silicon-photonics hardware that uses light, rather than only electrons, to move data and perform key computations. Its products are designed for hyperscale AI data centers that need to connect and coordinate thousands of accelerators efficiently as model sizes and cluster scales grow.

The company’s core portfolio spans photonic interconnects, co-packaged optics, and photonic AI accelerators. Lightmatter’s Passage platform uses 3D-stacked silicon photonics to provide extremely high-bandwidth links between chips, including the Passage M1000 interposer, which delivers a reported 114 Tbps of total optical bandwidth for advanced AI infrastructure designs. Its Envise photonic processor is intended to offload matrix operations using light while electronic components handle control and memory, targeting better performance-per-watt than conventional accelerators. Lightmatter works with foundry and packaging partners such as GlobalFoundries and ASE to make these systems deployable in production data centers.

In October 2024, Lightmatter raised a $400 million Series D round led by T. Rowe Price at a $4.4 billion valuation, bringing total capital raised to approximately $850 million. The company is headquartered in Mountain View, California, and has publicly discussed potential IPO timing as it scales manufacturing and deployment of its photonic hardware. This profile is based on publicly available information and is provided for informational purposes only. Summit Ventures is not affiliated with Lightmatter and does not offer or recommend securities; Summit Ventures facilitates introductions through a network of partners.